THREE MATCHED GROUPS, EACH CONSISTING OF 27 YOUNG ADULT MALES BETWEEN 17 AND 21 YEARS OF AGE, WERE SELECTED TO REPRESENT THE FOLLOWING CATEGORIES: THOSE ARRESTED AND CHARGED WITH AN ALCOHOL-RELATED OFFENSE, THOSE ARRESTED AND CHARGED WITH OTHER THAN AN ALCOHOL-RELATED OFFENSE, AND THOSE WITH NO KNOWN ARREST RECORD OR CHARGED OFFENSE OTHER THAN A MAXIMUM OF TWO MOVING TRAFFIC VIOLATIONS. THE ALCOHOL OFFENDER GROUP CONSISTED OF YOUNG ADULTS ARRESTED AND CHARGED DURING A 3-MONTH PERIOD IN THE CITIES OF KALAMAZOO AND BATTLE CREEK, MICH. NONALCOHOL OFFENDERS AND CONTROL SUBJECTS WERE FROM THE SAME GEORGAPHICAL AREA. ALL OFFENDERS WERE GIVEN THE MINNESOTA MULTIPHASIC PERSONALITY INVENTORY, THE BARRON EGO STRENGTH SCALE, A QUESTIONNAIRE CONCERNING PERSONAL DATA AND DRINKING BEHAVIOR, AND THE PARK PROBLEM DRINKING SCALE APPROXIMATELY 1 TO 3 MONTHS AFTER THEY WERE PLACED ON PROBATION. ALCOHOL OFFENDERS SHOWED A LOWER LEVEL OF MENTAL HEALTH, LOWER EGO STRENGTH SCALE SCORES, MORE EXCESSIVE DRINKING, AND GREATER PROBLEM DRINKING BEHAVIOR THAN CONTROLS. ALCOHOL AND NONALCOHOL OFFENDERS WERE SIMILAR IN MENTAL HEALTH AND EGO STRENGTH SCALE SCORES, BUT THE FORMER GROUP DRANK SIGNIFICANTLY MORE AND SHOWED MORE PROBLEM-DRINKING BEHAVIOR. THE IMPLICATIONS OF THE FINDINGS SUGGEST THAT EXCESSIVE DRINKING IN THE YOUNG ADULT OFFENDER POPULATION MAY BE LEARNED AND MAINTAINED BY PSYCHOLOGICAL, PHYSIOLOGICAL, AND/OR SOCIOLOGICAL REINFORCEMENT. SUPPORTING DATA AND REFERENCES ARE INCLUDED. (DEP)